Torrid (ticker: CURV) is an omni-channel plus-size apparel retailer targeting American women in their 20s and 30s. The company operates over 600 stores located primarily in malls throughout the US and Canada.

I expect CURV’s share price to rebound from the 4/6/2023 price of $4.25 to $10-20 by the end of 2023, with further appreciation beyond that level in 2024+ if a buyout doesn’t occur before then.

A more detailed downloadable slide deck PDF of this thesis with charts and photos is available at this dropbox link here (note that the slide deck was published on 3/28/2023): https://www.dropbox.com/s/08gbfxztod7uq8x/Torrid%20Investment%20Thesis%20%28full%20length%20version%29.pdf?dl=0

Investment thesis summary in 3 points

(1) Torrid’s stock price has fallen by nearly 90% since its 2021 post-IPO peak, but the issues that drove the stock price down are largely self-inflicted and are on the cusp of a reversal

• Torrid’s inability to adequately meet surging demand for its products in 2021 led it to overcompensate by ordering too much inventory. This inventory began to arrive at its stores just as inflation struck and demand softened starting in early 2022
• Management responded by heavily discounting products starting in mid-2022 to right-size inventory to meet current demand levels – this process finally came to an end a few months ago
• The negative margin impact of this inventory reduction will be transitory; growth and historical profitability should recover quickly starting in the next quarter or two

(2) Torrid’s business model should endure over time: a focus on fit over fashion addresses crucial plus-size customer pain points that other retailers ignore

• The offering resonates extremely well with customers, who respond through uniquely high levels of customer loyalty and repeat purchase behavior
• The company’s hybrid in-store and e-commerce model has a long history of excellent store-level economics and a healthy margin of safety on company-wide profit margins

(3) I expect the transitory margin and growth pressures to abate within a couple of quarters, leading to a normalization of the stock’s valuation multiple and a relatively quick rebound in share price

• The ultra-low valuation multiple of ~4x NTM maintenance free cash flow could enable the company to eventually repurchase large portions of the outstanding shares unless the stock price appreciates to higher levels quickly
• Investors can expect to make multiples of their money on this investment even in the absence of any future store growth or margin improvement beyond what should come from the near-term normalization of discounting and inventory alone
• Although a modest proportion of shares (1.2m of a total of 104 million) are sold short as of the last reporting date in mid-March, they represent a much more significant percentage of the public float. If the company’s outlook shows signs of improvement for the rest of 2023 during the upcoming May 2023 earnings call, this short interest could combine with a shareholder base that is reluctant to sell to create a potentially severe short squeeze

Relative valuation

Although I think Torrid’s intrinsic value will be what drives up the stock price up over time, it is worth considering relative valuation as a shorthand for demonstrating just how underpriced the stock is. Torrid is a superior business in comparison to apparel retailers such as Gap (GPS), American Eagle (AEO), and Urban Outfitters (URBN), yet trades at a large discount to those retailers on a go-forward FCF basis. While these retailers lack the plus-size specialization of Torrid, all four companies are roughly comparable omni-channel mall-based apparel retailers with a strong e-commerce presence. All four sell plenty of plus-size clothing.

To help normalize for recent NWC fluctuations and margin volatility for each competitor, I totaled up the cumulative FCF for each of these four companies since 2019 and compared their average annual FCF to each company’s respective current market cap. I found that CURV trades at 5x its average annual FCF over this four-year period, versus 15x for GPS, 21x for AEO, and 37x for URBN. CURV generated ~$350m of cumulative FCF over these four years, versus ~$600m for GPS, ~$500m for AEO, and ~$220m for URBN. Yet, those companies have an average market cap of $2.3bn, which is more than 5x that of Torrid.

Over this four year period, CURV grew revenue and EBITDA at a higher rate than these three competitors did, and reduced its share count by the most of the group (AEO grew diluted share count by 23%). Yet, CURV stock is down nearly 90% from its peak in August 2021 as compared to -64% for GPS, -60% for AEO, and -33% for URBN.

Torrid demonstrates capital efficiency that is in another league entirely when compared to these competitors. Over this four year period, Torrid’s average cash ROA was 14% vs 1% for GPS, 4% for AEO, and 2% for URBN. Similarly, Torrid’s average cash ROIC was 305% over this period vs 18% for GPS, 26% for AEO, and 11% for URBN.

CURV’s capital efficiency is largely derived from how much less it must invest in capex versus these competitors. Over the past four years, CURV has invested just 12% of Adjusted EBITDA into capex versus 61% for GPS, 41% for AEO, and 58% for URBN, all while growing more quickly than those competitors did. Cash conversion from EBITDA to FCF for CURV is also superior to that of competitors: since 2019, CURV converted 55% of Adjusted EBITDA to FCF versus 15% for GPS, 24% for AEO, and 15% for URBN.

Despite these stark differences in how EBITDA flows through into actual cash available to shareholders, many, if not all, sell side analysts use the average EBITDA multiple of this peer group to select a price target for CURV. It’s not just the sell-side that is using this misguided approach to valuation: many buy-side investors continue to value Torrid’s competitors on the basis of EBITDA multiples. For example, the ValueInvestorsClub buy thesis on AEO linked here relies solely on assumed EBITDA multiples to drive valuation.

In my view, this clear mispricing creates an opportunity for investors to acquire shares in an excellent business at a price that is unlikely to be this low for very long.

Simplified 24-month target price under different scenarios

In the interest of simplicity, all three cases shown here assume 3% annual comparable store sales growth and 35 new store additions per year.

Assumptions by case: Bear / Base / Bull

• NTM FCF less stock comp by 2025: $110m (bear) / $150m (base) / $190m (bull)
• Assumed exit multiple of NTM FCF less stock comp: 12x / 17x / 22x (note: this base case exit multiple is equivalent to 15x FCF if one doesn’t deduct stock comp)
• Cash generated and added to balance sheet by exit date: $140m / $180m / $220m
• Implied exit market cap: $1.5bn / $2.7bn / $4.4bn
• Current market cap: $440m
• MOIC assuming no buybacks or dividends occur: 3.4x / 6x / 10x
• Implied undiscounted share price in 2025: $14 / $26 / $42
• Implied discounted share price today: $12 / $22 / $35

Reasons for why Torrid shares are so mispriced

• Use of comparable company EBITDA multiples as a basis for valuation: In general, EBITDA should not be used to compare equity valuations of companies that have modest debt levels – especially given how different FCF as a % of EBITDA is between companies. Yet, these multiples are the key driver of valuation in almost all publicly available writing on CURV.
• Size: the small size of the public float inhibits many institutional investors from owning shares in sufficient quantities, so it is often just ignored completely.
• Volatility: Torrid’s small float and illiquidity create a setup for occasional bouts of extreme price volatility – it seems plausible that some more conservative investors shy away from even learning about Torrid as to not expose their portfolios to this level of price movement.
• Novelty: the company IPO’d less than two years ago and lacks an established investor base in comparison to retailers that have been public for decades.
• Investor lack of familiarity with Torrid: despite having a national store footprint, most public equities investors are firmly outside of Torrid’s target demographic and few have direct first-hand exposure to the company’s stores or brand.
• Temporary limit on repurchases of shares: I don’t think I’ve ever seen a situation before in which a company could realistically repurchase its entire public float in any given quarter, but that is where Torrid sits today. However, Sycamore (rightfully) refuses to sell any more of its stake at prices below $10++. So, until the price recovers enough for Sycamore to sell some of its stake, the company is likely limited in its ability to repurchase additional shares. A special dividend could be used as an alternative to address this unique situation.
• Mistaken treatment of lease liabilities as debt: Bloomberg and some other investment data services sometimes add retailers’ total lease liabilities (no matter the cost of those leases or whether they are cancellable or not) to be added to their calculation of “total debt” in a way that investors who don’t do the research themselves might be led to think that this company is much more levered than it is – and that the EV / EBITDA multiple is higher than it really is. For Torrid, such an assumption adds $220m of “debt” from leases, even though there is an offsetting right-of-use asset and many of the leases can be cancelled.
• Rates-driven market swings: the increase in rates and high inflation has led to many small growth companies (plus those that were recently growth companies until the past couple quarters) to be severely punished by the market to a degree that is much greater than what a 1-2% change in the long-term interest rate environment implies for these companies’ intrinsic value per share.

Torrid’s public float and trading volume

Torrid’s public float is small: Sycamore continues to own 79% of the outstanding shares. The number of shares owned by non-Sycamore entities is ~20 million. Even at its peak, the public float was worth just $400m. The size of the IPO was limited because Torrid had little use for incremental new capital – Sycamore sold just 12 million shares as part of the IPO and the company retained none of that cash.

Despite a small float, it is still possible to build a material and ultimately valuable position: excluding the elevated trade volumes during the first earnings report after the IPO, the 5 quarterly earnings leading up to the March 2023 report have generated trade volumes averaging 9 million shares in the two weeks following any given earnings announcement. This volume even exceeded the total share count of the public float in some quarters.

Corporate history

Torrid was launched in 2001 by the LA-based apparel retail chain Hot Topic. By 2008, revenue exceeded $150 million from approximately 125 stores. Hot Topic and Torrid were sold to Sycamore Partners for $600 million in 2013 in a take-private deal.

Sycamore Partners is a New York-based PE firm with $10+ billion of committed capital that specializes in retail investments. Sycamore’s returns have been spectacular: By 2017, Fund I had generated a 43% IRR after fees vs an industry average of 19% over that time period. Even with the multi-billion dollar mark-to-market loss on Torrid since its IPO, Fund I’s gross MOIC was still 2.0x with a net IRR of 25% through June 2022.

Torrid was formally spun-out from Hot Topic in 2015. The company first filed for an IPO in July 2017, but later that year retracted the plan to go public, citing market conditions. Torrid successfully completed an IPO in July 2021, selling 12.65 million shares at $21 per share (top of guided range). The standard 180 day lockup period following the IPO prevented Sycamore from selling additional shares until early 2022. By the time the lockup period had expired, shares had fallen to $10 – down more than 70% from its mid-$30s peak.

The plus-size women's apparel market

The plus-size women's apparel market is an $85 billion per year industry and accounts for just under half of the country’s total spend on apparel by women. Nearly 70% of America’s ~130 million women now qualify as plus-size (size 10+). Yet, of the 50,000 women’s specialty retail apparel stores in the US, under 2% are dedicated plus-sized stores. This equates to around 1 plus-sized store for every 50 mass-market store. Therefore, the vast majority of plus-size women are not purchasing from a dedicated plus-size retailer.

The plus-size portion of the market is growing at 2x the rate of the non-plus-size market, not only due to demographic change, but also due to this portion of the market being historically under-served. A 2016 study revealed that the average clothing size of American women increased by nearly one size every five years in the two decades preceding 2010, for a total increase from size 14 to size 16/18 over that period.

Apparel spend per person is materially lower for plus-size women than it is for non-plus-size women. A recent study commissioned by Torrid found that 78% of plus-size women would spend more on clothing if they had more options consistently available in their size and fit. To me, this suggests that Torrid’s addressable market is quite large and underserved.

Issues with plus-size shopping at mass-market retailers

Plus-size shoppers report numerous issues and problems with the retailing experience offered by mass-market apparel retailers. For example, most mass market retailers, brands, and clothing manufacturers simply design clothes for non-plus-size women before “grading up” the product to make it available to plus-size women – this approach causes serious problems on fit across different product lines. In addition, plus-size shoppers report an elevated frequency of out-of-stock sizes and as well as express disappointment over finding that certain products aren’t offered in plus-size at all by some mass-market retailers. Further customer frustration emanates from retailers’ habit of allocating just a section or a corner of a mass-market store towards the plus-size category in a way that effectively segregates these sizes from the rest of the store’s merchandising efforts.

These anecdotal challenges are supported by data: a 2017 UTNE survey of plus-size women found that only 18% of respondents could locate a plus-size mannequin at a mass market retailer selling plus-size clothing, 14% said they could easily locate images of people who looked like them in the store, and just 13% said they could see the store entrance from the plus-size section, implying it was typically hidden in the back of the store.

Overall, I think mass-market retailers will continue to improve their plus-size offerings in coming years, but will still fall far short in meeting customer needs as compared to a specialized offering such as Torrid’s.

Sizing varies significantly across brands, products, and time periods

Other plus-size apparel retailers offer clothing from multiple brands, which creates problems for customers as consistency in fit between brands is non-existent. A multi-decade shift to brand-centrism has allowed each brand to develop its own sizing system. Because Torrid is vertically integrated, it sells very few third party branded goods; other brands are generally only sold at Torrid in areas such as sunglasses or other accessories.

Fit and sizing have also steadily changed over time in a way that further complicates fit challenges. So-called “vanity sizing” allows clothing manufacturers to flatter customers by revising sizes downward

While vanity sizing is not a bad thing per se, this practice does still make the shopping process more difficult for those seeking consistency. Research published in The Washington Post last year illustrates how a size 8 dress today is about the same as a size 16 dress in 1958.

Furthermore, note that being able to find one’s correct size is a problem affecting women more so than men. This is because many men’s garments are sold by measurements (e.g. 34 inch arm and a 16.5 inch neck size on a dress shirt) as opposed to broader size categories (such as XL). Studies show that most women wear about three different sizes, while men typically go between two sizes. This difference is borne out in online return rates for women’s apparel versus men’s apparel. Women’s clothes are returned 50-75% more often than are men’s clothes. Torrid’s e-commerce return rate is just 9%, which is much lower than the ~20% industry average.

Torrid’s product offering and customer value proposition

Torrid employs in-house teams of specialized designers, artists, and product engineers to internally develop products that account for the vast majority of sales. Unlike other brands, Torrid does not rely on mannequins during the design and fit process, instead opting to fit all products on plus-size models, staff, and a select group of engaged customers.

Torrid uses a data-driven design process. The company has created a database of fit specifications based on years of testing, measuring, and cataloging over 13,000 garments each year on its models 1

Rather than having sizes range from 10-30 in a typical plus sizing approach, Torrid simplifies the experience by re-sizing all items down to a simpler size 00 (the smallest size, equivalent to a size 10) through size 6 (equivalent to a size 30).

Torrid has a vertically-integrated approach to sourcing and manufacturing that allows for the incorporation of sales data and customer feedback into the sizing of additional inventory purchases. Overall, my research suggests Torrid has been highly effective in leveraging its data advantage to build-out the industry’s best processes for creating plus-size apparel that meets its customers’ complex needs.

Torrid combines its deep online catalogue of apparel with a specialized in-store experience focused on fit. The in-store shopping experience at Torrid has received Net Promoter Scores of around 90 – I don’t know of many apparel retailers with such measurably high levels customer affinity.

Loyalty and connection with customers

My research has unveiled an unusually high level of brand loyalty for Torrid’s customers. This brand loyalty is borne out in the fact that in normal pre-pandemic times, 96% of revenue from ex